Cement & Concrete Composites 31 (2009) 107113

Contents lists available at ScienceDirect

Cement & Concrete Composites

journal homepage: www.elsevier.com/locate/cemconcomp

Properties of self-compacting concrete prepared with recycled glass aggregate

S.C. Kou, C.S. Poon *
Department of Civil and Structural Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China

a r t i c l e

i n f o

a b s t r a c t
The effects of recycled glass (RG) cullet on fresh and hardened properties of self-compacting concrete (SCC) were investigated. RG was used to replace river sand (in proportions of 10%, 20% and 30%), and 10 mm granite (5%, 10% and 15%) in making the SCC concrete mixes. Fly ash was used in the concrete mixes to suppress the potential alkali-silica reaction. The experimental results showed that the slump ow, blocking ratio, air content of the RGSCC mixes increased with increasing recycled glass content. The compressive strength, tensile splitting strength and static modulus of elasticity of the RGSCC mixes were decreased with an increase in recycled glass aggregate content. Moreover, the resistance to chloride ion penetration increased and the drying shrinkage of the RGSCC mixes decreased when the recycled glass content increased. The results showed that it is feasible to produce SCC with recycled glass cullet. 2008 Elsevier Ltd. All rights reserved.

Article history: Received 1 March 2008 Received in revised form 2 December 2008 Accepted 5 December 2008 Available online 14 December 2008 Keywords: Recycled glass Self-compacting concrete Property

1. Introduction A huge amount of solid waste is disposed of at the landlls in Hong Kong every year and the landll capacity is expected to be exhausted within 610 years time [1]. The reuse and recycling of solid waste is high on the agenda of sustainable waste management. Waste glass, especially waste glass beverage bottles, is a potential candidate for recycling [2]. Generally, in other parts of the world, it is known that most of recycled glasses collected, especially beverage bottles, are re-melted and used to reproduce new glass. However, not all the waste glasses are suitable for new glass production, because some waste glasses are contaminated with papers and other substances or colour mixed, which make these glasses not suitable for glass bottle re-production. Also, in Hong Kong there is currently no glass manufacturing industry. Hence, research is needed to nd other outlets for the waste glass, especially in the construction industry. A number of studies were conducted in the 1960s to try to use crushed waste glasses as aggregates for concrete [36]. In the past 10 years, the use of glass as concrete aggregates has again attracted much research interests due to high disposal costs for waste glasses and environmental regulations [713]. Meyer and Baxter [8,14] conducted very extensive laboratory studies on the use of crushed glasses as aggregates. They found that a practically feasible concrete mixture could be produced by using 100% crushed glass as aggregates, 80% ASTM Type III Portland cement and 20% metakaolin as cementitious materials and proper amount of superplasticizer. Chen et al. [13] investigated the properties of concrete containing various waste E-glass (glass derived from elec* Corresponding author. Tel.: +852 2766 6024, fax: +852 2334 6389. E-mail address: cecspoon@polyu.edu.hk (C.S. Poon). 0958-9465/$ - see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.cemconcomp.2008.12.002

tronic products) contents. Waste E-glass particles were obtained from electronic grade glass yarn scrap by grinding. The size distribution of cylindrical glass particle was from 38 to 300 lm and about 40% of E-glass particle was less than 150 lm. Based on the properties of hardened concrete, the optimum E-glass content was found to be 4050% by mass [13]. There are quite a number of applications of using recycled glass in the construction industry over the world. The application includes using glass in asphalt concrete (glassphalt), normal concrete, back-lling, sub-base, tiles, masonry blocks, paving blocks and other decorative purposes [1522]. But practical applications of using recycled glass in structural concrete are limited. A major concern regarding the use of glass in concrete is the chemical reaction that takes place between the silica-rich glass particles and the alkali in the pore solution of concrete, i.e., alkali-silica reaction (ASR). This reaction can be very detrimental to the stability of concrete, unless appropriate precautions are taken to minimize its effects. Such preventative actions could be achieved by incorporating a suitable pozzolanic material such as y ash, ground blast furnace slag or metakaolin in the concrete mix at appropriate proportions [23,24]. Self-compacting concrete (SCC) is considered as a concrete which can be placed and compacted under its self-weight with little or no vibration effort, and which is at the same time cohesive enough to be handled without segregation or bleeding. Previous investigations showed that the use of y ash and blast furnace slag in SCC reduced the dosage of superplasticizer needed to obtain similar slump ow compared to concrete made with Portland cement only [25]. Also, the use of y ash improved the rheological properties and reduced thermal cracking of the concrete produced [26]. Kim et al. [27] studied the properties of super-owing concrete containing y ash and reported that the replacement of

108

S.C. Kou, C.S. Poon / Cement & Concrete Composites 31 (2009) 107113

cement by 30% y ash resulted in excellent workability and owability. Other researchers [28] evaluated the inuence of supplementary cementitious materials on workability and concluded that the replacement of cement by 30% of y ash can signicantly improve rheological properties. Fly ash cannot only be used to improve the fresh and hardened properties of SCC, but it can also be used to suppress the expansion due to ASR in Portland cement concrete. The duo advantages of using y ash in SCC and recycled glass warrant a study to explore the feasible use of recycled glass aggregate in SCC which is a subject that has not been reported in the literature. 2. Experimental details 2.1. Materials 2.1.1. Cementitious materials ASTM Type I Portland cement with specic gravity of 3.15 and Blaine neness of 3519 cm2/g was used. A y ash that met the requirements of ASTM class F ash with a relatively low CaO content of <3%, Blain neness of 3960 cm2/g, and specic gravity of 2.31 was used. The physical properties and chemical compositions of the cement and class F y ash are presented in Table 1. 2.1.2. Aggregates Twenty millimetre maximum size crushed granite and river sand sourced from mainland China were used as natural coarse and ne aggregates, respectively. Table 2 presents their particle grading. The coarse and ne aggregates each had a specic gravity of 2.62 and water absorption of 0.86% and 0.87%, respectively. 10 mm and <5 mm recycled glass (RG) cullet were used in this study. The glass cullet was mainly waste beverage glass bottles and

was provided by the Environmental Protection Department (EPD) of Hong Kong through a contractor who carried out the collection, washing and crushing process. The RG was a blend of three different colors of glass bottles (30% colorless, 40% green and 30% brown). The mixed RG was used as aggregates for producing the mortar bars and SCC mixes in the experimental programme. The recycled glass sand (<5 mm) had the bulk specify gravity (SSD) of 2.49, neness modulus of 4.25 and a water absorption of 0.36. The grading of the RG aggregates satised the requirement for aggregates of BS 882: 1983 [29]. The particle size distribution of the RG is shown in Table 2. 2.1.3. Admixture In this study, a superplasticizer ADVA 109, based on the new comb polymer technology, was used in all concrete mixes. The ADVA 109 contained no added chloride and weighs approximately 1.045 0.02 kg/l. 2.2. Mix proportions 2.2.1. Mortar bar mixes The potential ASR expansion of the prepared mortar bars was assessed in accordance with ASTM C1260-02 [30]. Two series of mortar bars were prepared. In Series I, OPC was used as the cementitious material. Four mortar bar mixes, which used 0%, 15%, 30% and 45% RG sand to replace river sand respectively, were prepared. In Series II, y ash was used to replace 33% by weight of cement, and the RG sand replacement levels were the same as in Series I. The mix proportions are shown in Table 3. 2.2.2. Self-compacting concrete mixes A total of four SCC mixes were produced, and their detailed mix proportions are presented in Table 4. These included one control mix, and three mixes made with RG to replace 10mm granite (at 5%, 105 and 15% levels), and river sand (at 10%, 205 and 30% levels), respectively. For all SCC mixes, the amount of the cementitious materials used was generally maintained at approximately 500 kg/m3 of concrete and with a free water-to-binder ratio of 0.37. The ratio of ne-to-coarse aggregates was about 49:51. All the SCC mixes were designed to have a 28-day compressive strength of about 60 MPa. 2.3. Casting and curing of test specimens All the concrete mixes were mixed for 5 min in a laboratory pan mixer. Tests were conducted on fresh concrete mixes to determine slump ow, ow time, segregation resistance, L-box ratio, wet density and air content.

Table 1 Physical and chemical properties of cement and y ash. Chemical analyses (%) Calcium oxide (CaO) Silicon dioxide (SiO2) Aluminium oxide (Al2O3) Ferric oxide (Fe2O3) Magnesium oxide (MgO) Sodium oxide (Na2O) Potassium (K2O) Sulfur trioxide (SO3) Loss on ignition Physical properties Specic gravity Blaine neness (cm2/g) ASTM Type I cement 63.15 19.61 7.33 3.32 2.54 0.13 0.39 2.13 2.97 3.16 3519 ASTM class F y ash <3 56.79 28.21 5.31 5.21 0.45 1.34 0.68 3.90 2.31 3960

Table 2 Particle size distributions and density of aggregates. Sieve size (mm) 37.5 20 14 10 5 2.36 1.18 0.6 0.3 0.15 0.075 Relative density (g/cm3) River sand (% passing) 100 99 96 87 70 26 2 0.6 2.62 10 mm granite (% passing) 100 94 21 4 2.62 20 mm granite (% passing) 100 95 18 4 2.62 10 mm recycled glass (% passing) 100 99 18 2 2.49 Recycled glass sand (% passing) 100 99 64 35 20 10 5 1.4 2.49

S.C. Kou, C.S. Poon / Cement & Concrete Composites 31 (2009) 107113 Table 3 Mix proportions of mortar bars. Notations Series I Control G15 G30 G45 Series II Control-F G15-F G30-F G45-F RG (%) 0 15 30 45 0 15 30 45 Cement (g) 440 440 440 440 440 440 440 440 RG (g) 148.5 297 445.5 126.5 253 380 Sand (g) 990 751.5 693 544.5 844.5 718 591.5 464.5 Fly ash (g) 145.5 145.5 145.5 145.5

109

the sieve was measured and expressed as a percentage of the weight of the original sample on the sieve. 2.4.1.3. Blocking ratio (L-box test). The L-box test was performed in according with EFNARC standards [31]. This test has been used to assess the owability and passing ability of concrete. During the test, SCC was allowed to ow upon the release of a trap door from the vertical section to the horizontal section via a few reinforcement bars of a L-shape box. The height of the concrete at the end of the horizontal section was compared to the height of concrete remained in the vertical section. The ratio is usually between 0.8 and 1.0. If the ratio is less than 0.8, the viscosity of the SCC mixes is too high which would cause blockage around the reinforcement. 2.4.1.4. Air content test. The air content of the RGSCC was determined in accordance with BS 1881 Part 114 [32]. 2.4.2. Determination of the hardened properties of RGSCC 2.4.2.1. Compressive and splitting tensile strengths. The compressive and splitting tensile strengths of RGSCC were measured using a Denison compression machine with a loading capacity of 3000 kN. The loading rates applied for the compressive and tensile splitting tests were 200 kN/min and 57 kN/min, respectively. The compressive strength test was carried out at the ages of 1, 4, 7, 28 and 90 days while the tensile splitting strength test was carried out at the ages of 28 and 90 days. 2.4.2.2. Static modulus of elasticity tests. Static modulus of elasticity of the RGSCC was determined on three 100 200 mm cylindrical specimens according to ASTM C 469-02 [33]. 2.4.2.3. Drying shrinkage test. A modied British Standard Test Method (BS1881, Part 5: 1970) was used for the drying shrinkage test. For the specimens (70 70 285 mm) were prepared for the drying shrinkage test, after removing them from the curing tank after 28-day of curing, the initial length of each specimen was measured. After the initial reading, the specimens were conveyed to a drying chamber with a temperature of 55 C and a relative humidity of 95% until further measurements at 1, 4, 7, 28, 56, 90 and 112 days after the initial measurement. Before each measurement was taken on the scheduled day, the specimens were rst removed from the drying chamber and conveyed to a cooling chamber for about 4 h in order to cool the specimens to a constant temperature of 25 C and a relative humidity of 75%. The length of each specimen was then measured within 15 min before delivering the specimens back to the drying chamber for the subsequence drying process. The procedure of drying, cooling and measuring continued until the nal length measurement was recorded at 112 days. 2.4.2.4. Determination of chloride ion penetrability. The chloride ion penetrability was determined in accordance with ASTM C1202-97 [34] using 50 mm thick concrete discs obtained from the 100 200 mm concrete cylinders. The resistance of concrete to chloride ion penetration was represented by the total charge passed in coulombs during a test period of 6 h. In this study, the

From each concrete mix, fteen 100 mm cubes were cast for the determination of compressive strength. Twenty eighty 100-mm diameter 200 mm cylinders were cast for the determination of splitting tensile strength, static modulus of elasticity and the resistance to chloride ion penetration. To assess the self-compactability of these mixes, half of the cylinders were cast as SCC (with no vibration/compaction applied), while the other half were cast with a vibration/compaction process similar to that of normal concrete (Nml). Six 70 70 285 mm prisms were cast for determining the drying shrinkage and alkali-aggregate reaction (ASR). After casting, all the cast specimens were covered with plastic sheets and water-saturated burlap and left in the laboratory at 20 3 C for 24 h. The specimens were then demoulded and transferred to a standard water curing tank at 27 C until the age of testing. On the other hand, the mortar bars were cast according to ASTM C1260 using a water to cement ratio of 0.47. The sieved aggregates were mixed with the cementitious materials at the stipulated mix proportions. Then, the mixes were cast into steel moulds which had internal dimensions of 25 25 285 mm. 2.4. Testing of concrete mixes 2.4.1. Determination of the fresh properties of RGSCC 2.4.1.1. Slump ow test. The slump ow test was used to evaluate the free deformability and owability of the SCC in the absence of any obstruction. A standard slump cone was used for the test and the concrete was poured in the cone without compaction. The slump ow value is represented by the mean diameter (measured in two perpendicular directions) of concrete after lifting the standard slump cone. The measurements were repeated at 30, 60, 90 and 120 min after the initial mixing to evaluate the slump loss properties of the SCC mixes. 2.4.1.2. Segregation test. The GTM screen stability test method, developed by the French contractor (GTM) [31] was adopted to assess the segregation resistance of the fresh SCC mixes. The method consisted of taking 10 l of concrete and allowing the concrete to stand for 15 min in a bucket covered with a lid to prevent evaporation. After that, half of the concrete was poured onto a 5 mm sieve of 350 mm diameter, which sat on a sieve pan on a weigh scale. After two minutes, the mass of mortar which passed through

Table 4 Mix proportions of SCC (kg/m3). Notation Control G15 G30 G45 OPC 375 375 375 375 PFA 125 125 125 125 20 mm 422 422 422 422 10 mm 400 380 360 340 Sand 860 774 688 602 10 mm Glass 0 81 162 246 Glass sand 0 19 38 54 Water 185 185 185 185 SP (l/m3) 8.5 8.2 8.0 7.8 W/C 0.37 0.37 0.37 0.37

110

S.C. Kou, C.S. Poon / Cement & Concrete Composites 31 (2009) 107113

chloride ion penetrability test was carried out on the concrete specimens at the ages of 28 and 90 days. 2.4.2.5. Alkali-silica reaction. The accelerated mortar bar test was carried out in accordance with ASTM C1260 (80 C, 1 N NaOH solution) but the duration of the ASR test was extended to 28-days (the standard test period is 14 days). The expansion of the mortar bars were measured at 1, 4, 7, 14 and 28 tested ages. 3. Results and discussion 3.1. Fresh properties The test results of the control and RGSCC mixes are summarized in Table 5. Table 6 lists some typical acceptance criteria for SCC in Europe [31]. Comparing the results with the European criteria, it can be seen that all RGSCC mixes showed satisfactory properties. 3.1.1. Slump ow and blocking ratio (L-box test) It can be seen that the initial slump ow of RGSCC mixes was similar to the control although the dosage of superplasticizer was decreased (see Table 3). This is attributed to the weaker cohesion between the glass aggregates and the cement paste due to their smooth and impermeable surfaces. This result is similar to that of Terro [35] who reported that the slump of normal concrete was the same as that of concrete with recycled glass cullet. The changes in slump ow of the concrete mixes with time are shown in Fig. 1. The owability decreased with time due to the hydration of cement, which increased the cohesiveness of the cement paste. The rate of slump loss decreased with the increase in RG aggregate content. All RGSCC mixes had lower rates of slump loss than the control. As shown in Table 5, the blocking ratios varied from 0.84 to 0.87 for the RGSCC mixes. The results indicate that the RGSCC mixes prepared in this study achieved adequate passing ability and maintained sufcient resistance to segregation around congested reinforcement areas. According to the European specication [31] the blocking ratio of SCC should be between 0.8 and 1.0. 3.1.2. Segregation ratio The segregation ratios for the control and RGSCC mixes are shown in Table 5. It can be seen that the segregation ratio of the SCC tended to increase with an increase in the RG aggregate con-

800 750 Slump flow (mm) 700 650 600 550 500 0 30 60 Time (min)
Fig. 1. Change of slump ow with time.

Control RG15 RG30 RG45

90

120

tent. The resistance to segregation of RGSCC mixes was lower than the control. The segregation ratio varied from 10.6% to 12.3%. The European specication [31] suggests that if the segregation ratio is between 5% and 15%, the segregation resistance of the SCC is considered satisfactory. If the segregation ratio is below 5%, the resistance is deemed excessive and is likely to affect the surface nish. On the other hand, if the segregation ratio is above 15%, and particularly above 30%, there is a strong likelihood of segregation. The results indicated that the RGSCC mixes prepared in this study achieved adequate segregation resistance.

Table 7 Compressive strength, splitting tensile strength and density of SCC. Notation Compressive strength (MPa) 1-day Control G15 G30 G45 23.5 (0.2) 22.3 (0.1) 20.8 (0.2) 19.2 (0.1) 3-day 36.6 (0.3) 35.8 (0.3) 33.6 (0.4) 31.2 (0.2) 7-day 47.3 (0.1) 46.6 (0.2) 44.4 (0.3) 42.3 (0.4) 28day 67.4 (0.2) 66.4 (0.4) 64.6 (0.2) 61.7 (0.3) 90day 75.4 (0.2) 74.9 (0.2) 72.9 (0.6) 68.5 (0.5) Splitting tensile strength (MPa) 28-day 3.42 (0.03) 3.26 (0.04) 3.19 (0.03) 3.03 (0.03) 90-day 4.18 (0.02) 4.02 (0.03) 3.90 (0.03) 3.71 (0.01) 2378 (8) 2370 (6) 2337 (5) 2285 (4) Density (kg/m3)

Table 5 Fresh properties of SCC. Notation Slump ow Final time (s) Control G15 G30 G45 42.5 42 43 41 Final diameter (mm) 755 750 760 750 L-box ratio Final time (s) 36 40 39 38 Ratio Air content (%) Segregation ratio (%) Wet density (kg/m3)

90 80 Compressive strength (MPa)

0.88 0.87 0.86 0.84

3.1 2.6 2.7 2.9

10.4 10.6 11.5 12.3

2290 2280 2260 2250

70 60 50 40 30 20 10 0 0 7 14 21 28 35 42 49 56 63 70 77 84 91 98 Time (day)
Fig. 2. Development of compressive strength of SCC.

Table 6 General acceptance criteria of SCC [31]. Test method Unit Typical range of values Minimum Slump ow GTM screen stability test L-box, H2/H1 mm % Ratio 650 0 0.8 Maximum 800 15 1.0

Control RG15 RG30 RG45

S.C. Kou, C.S. Poon / Cement & Concrete Composites 31 (2009) 107113

111

3.2. Hardened properties 3.2.1. Compressive, tensile splitting strengths and density The test results of the compressive strength of the control and RGSCC mixes are summarized in Table 7 and the development of compressive strength of the concrete are shown in Fig. 2. Each

Table 8 Static elasticity modulus and chloride ion penetration of SCC. Notation Static elasticity modulus (GPa) 28-day Control G15 G30 G45 29.8 29.2 28.3 27.2 (0.6) (0.5) (0.4) (0.4) 90-day 32.9 31.4 30.2 29.3 (0.4) (0.3) (0.3) (0.3) Chloride ion penetration (coulombs) 28-day 442 373 228 176 (5) (4) (5) (5) 90-day 107 108 108 100 (4) (3) (4) (4)

presented value is the average of three measurements. It is evident in Table 7 and Fig. 2 that at all test ages, the use of RG as sand and 10 mm coarse aggregate replacement decreased the compressive strength of the SCC mixes compared with the control. The corresponding reduction in the 28-day strength for RG15, RG30 and RG 45 was 1.5%, 4.2% and 8.5%, respectively. These results are similar to that of Park et al. [36] who reported that the compressive strength of normal concrete was only slightly less than the strength of concrete made with recycled glass. This may be attrib-

0.3 0.25 Expansion (%) 0.2 0.15 0.1 0.05 0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Time (days)
Fig. 5. Results of ASR expansion of the mortar bars prepared in Series I.

Control RG15 RG30 RG45

31 28-day elastic modulus (GPa) 30 29 28 27 26

0.565

y = 2.9858x
2

R = 0.8263

0.3
25
60 62 64 66 68 70

0.25 Expansion (%) 0.2 0.15 0.1 0.05 0 0 2 4

28-day compressive strength (MPa)

Fig. 3. Relationship between 28-day compressive strength and 28-day elastic modulus of SCC.

Control-F RG15-F RG30-F RG45-F

Table 9 Drying shrinkage of SCC at age of 112 days. Mix notation Control G-15 G-30 G-45 Recycled glass (%) 0 15 30 45 Drying shrinkage (Length change %) 0.038 0.036 0.034 0.031 (0.003) (0.004) (0.004) (0.003)

10 12 14 16 18 20 22 24 26 28 30 Time (days)

Fig. 6. Results of ASR expansion of the mortar bars prepared in Series II.

0.04
0.3

0.035 0.03 Length change (%) 0.025 0.02 0.015 0.01 0.005 0 0 14 28 42 56 70 Time (day) 84 98 112
0 0 2 4 Expansion (%) 0.25 0.2 0.15 0.1 0.05

Control RG15 RG30 RG45

Control RG15 RG30 RG45

10

12 14

16

18

20

22 24

26

28

Time (day)
Fig. 7. Results of ASR expansion of SCC with varying percentage of RG.

Fig. 4. Drying shrinkage curves of SCC with varying percentage of RG.

112

S.C. Kou, C.S. Poon / Cement & Concrete Composites 31 (2009) 107113

Table 10 Relative properties of SCC and normally compacted mixes (SCC/Nml). Notation Splitting tensile strength 28-day Control G-15 G-30 G-45 0.97(0.02) 0.97(0.02) 0.98(0.02) 0.98(0.02) 90-day 0.99(0.01) 0.97(0.01) 0.97(0.01) 0.98(0.01) Static elasticity modulus 28-day 0.97(0.03) 0.99(0.02) 0.99(0.02) 0.99(0.02) 90-day 0.99(0.02) 0.99(0.01) 0.98(0.01) 0.98(0.02) Chloride ion penetration 28-day 1.09(0.02) 1.09(0.02) 1.05(0.02) 1.03(0.01) 90-day 1.07(0.01) 1.08(0.01) 1.08(0.01) 1.01(0.01)

uted to the decrease in bond strength between the cement paste and the RG, and the increase in FM of the ne aggregate which decreased the overall density of the concrete produced [37]. The results of the tensile splitting strength of the SCC mixes are presented in Table 7. Each presented value is the average of three measurements. The results indicate that the 28 and 90 days tensile splitting strength of the RGSCC mixes were lower than that of the control. The RGSCC mixes with 45% RG glass had the lowest tensile splitting strength. It is evident from Table 7 that the density of the RGSCC mixes decreased with the increase in RG content. This is mainly attributed to the difference in densities between the recycled glasses, natural sand and granite. 3.2.2. Static modulus of elasticity The results of the static modulus of elasticity are presented in Table 8. Each presented value is the average of three measurements. The elastic modulus decreased with the increase in RG content. Fig. 3 shows the relationship between compressive strength and modulus of elasticity at 28-days. The results show that there is a good correlation between the compressive strength and the modulus of elasticity for the prepared SCC mixes in this study. 3.2.3. Chloride ion penetration The results of the chloride ion penetration are presented in Table 8. The values are the average of two measurements. The chloride ion penetrability decreased with the increase in RG content which means the resistance to chloride ion penetration increased with an increase in recycled glass content. This was due to the recycled glass cullet had lower porosity than that of river sand and natural granite. Moreover, the RG sand which had a ner particle size distribution than that of river sand might have also resulted in a better packing efciency of the concrete at the ne scales. 3.2.4. Drying shrinkage The drying shrinkage values measured as function of time up to 112 days are shown in Table 9 and Fig. 4. Each presented value is the average of three measurements. It can be seen that the drying shrinkage values decreased with an increase in recycled glass content probably due to the lower water absorption characteristics of glass cullet. The drying shrinkage of the concrete mixes was well below the limit of 0.075% at 56 days specied by the Australian Standard AS 3600 [38]. 3.2.5. Alkali-silica reaction expansion The results of the mortar bars test in Series I and II are shown in Figs. 5 and 6, respectively. Fig. 5 indicate that although the ASR expansion was able to meet the requirements prescribed in ASTM C 1260 (<0.1% within 14 days) when the mortar bars were prepared with 45% RG, serious expansion and cracks were observed at 28-days. In comparison, the expansion of the control mortar bars in which only river sand was used as aggregates was minimal. The results in Fig. 6 show that all the ASR expansion of all the specimens were signicantly reduced by the use of y ash in the mortar mixes.

The ASR expansion rate of the RGSCC concrete mixes were measured using the concrete prisms prepared using the same exposure condition as the mortar bar test method (ASTM C 1260). The results (Fig. 7) show that the concrete prisms displayed extremely small expansion even with a RG content of 45% showing the expansion was suppressed by the use of y ash in the concrete mixes. All 28-day measurements showed less than 0.1% expansion. 3.2.6. Relative properties of SCC and normally compacted mixes The relative properties of the SCC to that of the normally compacted mixes are shown in Table 10. The data shows the relative ratios were close to one demonstrating the produced concrete mixes were truly SCC and as the hardened properties of the mixes with and without compaction were similar. Ho et al. [39,40] suggested that if the hardened properties ratio of SCC/Nml is between 0.95 and 1.05, the SCC is considered truly SCC. 4. Conclusions This paper has presented the experimental results of a study on the feasible use of recycled glass cullet for the production of SCC. Based on the results of this study, the following conclusions can be drawn: 1. The slump ow, blocking ratio, air content of the RGSCC mixes increased with increase in recycled glass content. The initial slump ows of all the SCC mixes prepared in this study were at least 750 mm. The blocking ratios varied from 0.84 to 0.88. 2. The compressive strength, tensile splitting strength and static modulus of elasticity of the RGSCC mixes decreased with an increase in recycled glass content. 3. The chloride ion penetrability of the RGSCC mixes was significant decreased when compared with the control. The resistance to chloride ion penetration increased when the recycled glass content increased. 4. The drying shrinkage of the RGSCC mixes decreased with an increase in the recycled glass aggregate content. 5. The overall assessment of both the fresh and hardened properties indicates it is feasible to produce SCC with recycled glass cullet. 6. The ASR expansion of all the specimens was signicantly reduced by the use of y ash. Acknowledgements The authors would like to thank the Hong Kong Polytechnic University for funding support and the Environmental Protection Department for providing the waste glass bottles. References
[1] Environmental Protection Department. Future landll development in Hong Kong. The Government Printer; 2006. [2] Environmental Protection Department. Monitoring of solid waste in 2005, Hong Kong. The Government Printer; 2006. [3] Pike RG, Hubbard D, Newman ES. Binary silicate glasses in the study of alkaliaggregate reaction. High Res Board Bull 1960;275:3944.

S.C. Kou, C.S. Poon / Cement & Concrete Composites 31 (2009) 107113 [4] Schmidt A, Saia WHF. Alkali-aggregate reaction tests on glass used for exposed aggregate wall panel work. ACI Mater J 1963;60:12356. [5] Phillips JC, Chan DS, Keller GW. Reuse glass aggregate in Portland cement. In: Proceedings of 5th mineral waste utilization symposium. Chicago, III: IIT Research Institute; 1972. p. 38590. [6] Johnston CD. Waste glass as coarse aggregate for concrete. J Test Eval 1974;2(5):34450. [7] Meyer C, Baxter S, Jin W. Alkali-aggregate reaction in concrete with waste glass as aggregate. In: Proceedings of 4th materials engineering conference: materials for the new millennium. Reston, VA.: ASCE; 1996. p. 138897. [8] Meyer C, Baxter S. Use of recycled glass and y ash for precast concrete. Rep. NYSERDA 98-18 (4292-IABR-IA-96) to New York State Energy Research and Development Authority, Dept of Civil Engrg and Engrg Mech. Columbia University, New York; 1998. [9] Pollery C, Cramer SM, De La Cruz RV. Potential for using waste glass in Portland cement concrete. J Mater Civil Eng 1998;10(4):2109. [10] Bazant ZP, Zi G, Meyer C. Fracture mechanics of ASR in concretes with waste glass particles of different sizes. J Eng Mech 2000;126:22632. [11] Byars EA, Morales-Hernandez B, Zhu HY. Waste glass as concrete aggregate and pozzolan. Concrete 2004;38(1):414. [12] Topcu IB, Canbaz M. Properties of concrete containing waste glass. Cement Concr Res 2004;34:26774. [13] Chen CH, Huang R, Wu JK, Yang CC. Waste E-glass particles used in cementitious mixtures. Cement Concr Res 2006;36:44956. [14] Meyer C, Baxter S. Use of recycled glass for concrete masonry blocks nal report 97-15. Albany, New York: New York State Energy Research and Development Authority; 1997. [15] Dhir RK, Limbachiya MC, Dyer TD. Recycling and reuse of glass cullet. Thomas Telford; 2001. [16] Clean Washington Center. Recycled glass in Portland cement concrete, November 1996. [17] CSIRO. Making concrete with glass now possible, 13 May 2002. [18] Sagoe-Crentsil K, Brown T, Taylor A. Guide specication recycled glass as sand replacement in premix concrete, CSIRO, June 2001. [19] Meyer C. Development of glass concrete products. Final report to New York Empire state development, Albany, New York, March 1999. [20] Robert J, Kirby. Final report: an investigation of performance of paving units made from recycled glass. National Center for Environmental Research; 2001. [21] Clean Washington Center. Best practices in glass recycling, November 1996. [22] CSIRO. Recycled Glass as Concrete Aggregate, April 2002. [23] Shayan A, Xu A. Value-added utilization of waste glass in concrete. Cement Concr Res 2004;34(1):819. [24] Lam CS, Poon CS, Chan D. Enhancing the performance of pre-cast concrete blocks by incorporating waste glass ASR consideration. Cement Concr Compos 2007;29(8):61625.

113

[25] Yahia A, Tanimura M, Shimabukuro A, Shimoyama Y. Effect of rheological parameters on self compactability of concrete containing various mineral admixtures. In: Skarendahl A, Petersson O, editors. Proceedings of the rst RILEM international symposium on self-compacting concrete. Stockholm; September; 1999. p. 52335. [26] Kurita M, Nomura T. Highly-owable steel ber-reinforced concrete containing y ash. In: Malhotra VM, editor. Am Concr Inst SP 178; 1998. p. 15975 [June]. [27] Kim JK, Han SH, Park YD, Noh JH, Park CL, Kwon YH, et al. Experimental research on the material properties of super owing concrete. In: Bartos PJM, Marrs DL, Cleland DJ, editors. Production methods and workability of concrete. E&FN Spon; 1996. p. 27184. [28] Miura N, Takeda N, Chikamatsu R, Sogo S. Application of super workable concrete to reinforced concrete structures with difcult construction conditions. Proc ACI SP 140. p. 16386. [29] BS 882. Specication for aggregates from natural sources for concrete. British Standards Institution; 1992. [30] ASTM C1206. Standard test method for plutonium by iron (II)/ chromium (VI) amperometric titration. American Society of Testing Materials; 2002. [31] EFNARC. Specication and guidelines for self-compacting concrete. European association for producers and applicators of specialist building products (EFNARC); 2002. [32] BS 1881-114. Test concrete part 114: methods for determination of density of hardened concrete. British Standards Institution; 1983. [33] ASTM C 469. Standard test method for static modulus of elasticity and Poissons ratio of concrete in compression. American Society of Testing Materials; 2002. [34] ASTM C 1202. Standard test method for electrical indication of concretes ability to resist chloride ion penetration. American Society of Testing Materials; 1997. [35] Mohamad JT. Properties of concrete made with recycled crushed glass at elevated temperatures. Build Environ 2006;41(5):6339. [36] Park SB, Lee BC, Kim JH. Studies on mechanical properties of concrete containing waste glass aggregate. Cement Concr Res 2004;34(12):21819. [37] Roberts LR. Microsilica in concrete, materials science of concrete. Westerville (OH): American Ceramic Society; 1989. p. 197222. [38] AS 3600. Concrete structures-incorporating AMD 1: May 2002 and AMD 2: October 2004. Australian Standard. [39] Ho DWS, Sheinn AMM, Ng CC, Lim WB Tam CT. Self-compacting concrete for Singapore. In: Proc 26th conference on our world in concrete and structures, Singapore. Singapore: CI-Premier Pte Ltd. 2001. p. 2939. [40] Ho DWS, Tam CT, Sheinn AMM. Some major issues of self-compacting concrete. Singapore: Construction and Housing Development Board; 2001. pp. 7481.